The existence of long-lasting cellular reservoirs of HIV-1 is one of the major hurdles in developing effective anti-retroviral therapies. These latently infected cells and tissues efficiently evade immune responses and remain dormant until activated, upon which they can generate a productive HIV-1 infection. This classic scenario of viral latency becomes even more difficult to study and model due to the extreme complexity of translating in vivo virus-cell interactions into a controlled in vitro system. The recent developments and constant improvements upon hematopoietic engraftment of human cells and tissues onto recipient immunocompromised murine scaffolds have made it possible to model complex human innate and adaptive immune responses in a small animal model. Specifically, HIV-1 infection has been successfully modeled in these humanized mice to mimic transmission, pathogenesis, host immune responses, and treatment. Here, we review the complexities surrounding modeling HIV-1 latency in vitro and in vivo and highlight the most recent humanized mouse models that support retroviral infection.